Author: Pete Gutteridge-Paye, VP of Delivery, Namos Solutions
There’s a familiar feeling in HR leadership circles right now: excitement, pressure, and a quiet fear of falling behind.
At a recent cross‑industry roundtable dinner, one theme cut through the noise: the speed of AI is overwhelming many organisations, and the worry isn’t simply “we don’t have AI”, it’s “how do we catch up when our foundations aren’t ready?”
What followed was one of the most honest conversations I’ve heard in a while: not about shiny AI features, but about the real blockers, such as data, integration, process design, trust, and change.
Here are the key takeaways, reframed for anyone who wasn’t in the room.
1) Legacy and fragmentation are the real “AI readiness” problem
A consistent challenge raised was the reality of legacy HR landscapes, systems that are either not integrated or integrated in ways that still don’t produce a single source of truth. Even where organisations have “consolidated,” the question remains:
- “We have separate but consolidated systems. With AI coming in, where do we go from here?”
- “We have one system, but is it ready for AI? Where do we start?”
The message was clear: AI doesn’t eliminate complexity. It exposes it. If foundational HR data is inconsistent, duplicated, or split across tools, AI can only work with what it can access, and what it can access may not be reliable.
2) AI won’t fix process gaps we chose to live with
A really important point came up: in many organisations, there are gaps in system delivery not because of technology limitations, but because of:
- human interactions and workarounds becoming “the process,” or
- deliberate choices not to implement something to fill the gap.
The group’s view was pragmatic: AI can’t fix all gaps if the organisation never fully defined or implemented the end‑to‑end process in the first place. In other words, if your HR operating model is patchy, AI becomes a very expensive way of automating inconsistency.
3) Heavy customisation creates an “AI tax”
Highly customised systems were a major concern. The discussion wasn’t anti‑customisation, it was about realism: how do updates and AI keep up (or even work) in environments where customisations are still being implemented or heavily relied upon?
This becomes the “AI tax”:
- higher maintenance burden
- slower adoption of new capabilities
- more integration complexity
- more uncertainty about what’s supported and what’s stable
If AI is the destination, customisation choices need to be re‑examined through a new lens: does this genuinely differentiate us, or does it slow the whole organisation down?
4) “Broken HR” often starts at day one: the wrong people, too late
One of the strongest insights was that broken HR processes are frequently a delivery/design problem, specifically who is involved at the start of an implementation. A clear view emerged: when the right people don’t design the solution early, organisations pay for it later through:
- rework in late project stages, and/or
- rework after go‑live when it’s most painful.
This matters for AI because AI amplifies whatever you build. If the solution design is compromised, AI won’t “repair” it, it will scale the consequences.
5) Employee experience vs business value: the tension AI exposes
A vibrant part of the conversation focused on the trade‑off between employee experience and business need/value. In some organisations, multiple systems exist specifically to meet employee experience needs, sometimes at the detriment of business objectives. But that comes at a cost: multiple systems mean a complex integration layer and that’s exactly what AI must navigate to retrieve context and make good decisions.
If experience and value aren’t aligned, AI doesn’t magically reconcile them, it forces leadership teams to confront the trade-offs.
6) Trust, adoption, and “tacit knowledge” are the hard parts
The roundtable returned again and again to human realities:
Tacit knowledge
A crucial question was raised: organisational memory lives in people’s heads. How does tacit knowledge get passed to a machine, and does it need to? If AI makes decisions solely on what it can access, then there’s a risk: it may miss the hidden context that humans rely on every day.
Generational change
There was strong discussion about how generations may experience AI differently, and what happens when people no longer have the option to use “human” channels in parallel with “machine” channels.
Chatbots and trust
Chatbots triggered a deeper point: trust in the answer and trust in whether it works at all. Even if the tech is available, will people use it?
And that links to a simple behavioural truth raised in the room: If users don’t see value, they won’t engage. —They’ll use old channels as long as those channels remain open (unless there’s a deliberate move to “AI first”).
The questions leaders are quietly asking (and should ask out loud)
Some of the most powerful moments of the evening were the questions, not the answers:
- Will we ever achieve a steady state? Or will we always feel behind the “next best thing”?
- Are we looking at the problem from the wrong angle, expecting AI to fit into HR, when perhaps humans need to assimilate to AI possibilities?
- If transactions are removed, what is HR’s value to the business, and what work is left for the workforce?
- Is the workplace ready for “AI-first,” or is this actually a generation shift requiring the right leaders at the right moment?
And perhaps the most provocative phrase of the night which I really cannot take credit for was: “Human is the new luxury.”
Practical takeaways: how to move forward without being overwhelmed
The dinner didn’t end with panic, it ended with pragmatism.
1) Think “marginal gains,” not big-bang
Do we really need to do everything at once? Or is AI deployment a structured activity that builds confidence and acceptance step-by-step?
2) Treat AI as a destination and design the route
AI is here, but it’s also a destination. The key question is: how do we arrive (even if we never fully “arrive”)? Knowing the destination gives direction, even if it shifts.
3) Align employee experience with organisational strategy
If employees happily use AI outside work, why is AI inside work met with greater resistance? The difference is often trust, usefulness, and how change is introduced, not the existence of AI itself.
A simple “AI readiness” checklist for HR leaders
If you’re feeling the pressure, here’s a practical way to frame next steps (based on the themes above):
- Single source of truth: do we trust our workforce data end-to-end?
- Integration reality: can AI access the right information without navigating an unstable integration maze?
- Process gaps: are we trying to automate decisions where the process is still undefined or workaround-driven?
- Design governance: did the right people shape the solution early enough to avoid rework?
- Trust & adoption: what makes users choose AI, and what legacy channels need redesigning or retiring?
Closing Thought
The strongest message I took away is this:
If HR is fragmented, AI won’t fix it. But clean data and integrated HCM can unlock real-time value and make AI adoption make sense.
If you’re wrestling with AI readiness, you’re not alone, and it isn’t isolated to HR. It’s shared across industries and across functions.
The winners won’t be the organisations that “buy AI first.” They’ll be the ones that build the foundations and deploy AI in a way humans can trust, adopt, and benefit from.
